← all companies · 2026-07-07 · moat: narrow · raw .md

Wipro: An Invoice, Dissected

One line on a piece of paper

Somewhere in a Fortune 500 accounts department, an invoice arrives from Wipro Limited. Strip away the logo and the legal boilerplate and one line carries the entire business model of Indian IT:

Software engineer, offshore, 160 hours — billed at so-many dollars per hour.

Take a scalpel to that line. The client pays, say, some amount X for the month. The engineer in Bengaluru who did the work is paid a fraction of it — call it X divided by n, where n has historically been a large and comfortable number. The gap between X and X/n is not all profit; it has many mouths to feed. Some of it pays for the "bench" — trained engineers currently between projects, kept warm like reserve players, because a services firm that cannot staff a contract next Monday loses the contract. Some pays for the buildings, the training, the managers of managers. A large slice pays for the sales machine — the people who spend eighteen months courting a client for a five-year deal. What remains, after every mouth, is the operating margin. At Wipro, last year, that remainder was 19 paise of every rupee billed.

That single line, multiplied by a few hundred thousand engineers, is a ₹92,624-crore-a-year business. It is also — and this is the puzzle at the heart of this chapter — the identical line that appears on invoices from TCS, which turns the same trade into returns on owners' capital three times higher. Same country, same talent pool, same clients, same decade. Different outcome. Before we solve that puzzle, we should understand why the gap between X and X/n existed at all.

The oldest trade, in its newest costume

Economics has one relentless law: identical things should not sell for different prices in different places, because someone will buy cheap and sell dear until the gap closes. That someone is called an arbitrageur. For most of history, goods could make the journey but labour largely could not — a skilled worker's price stayed local because the worker had to live where the work was.

Software broke that law's enforcement mechanism. Code is pure information; it travels at the speed of light and arrives unchanged. The moment intercontinental bandwidth became cheap — the late 1990s — an engineer in Bengaluru became, for the first time in economic history, functionally present in an office in Dallas. The wage gap between the two cities, previously protected by oceans, was suddenly harvestable at industrial scale. Indian IT services is that harvest: buy engineering capacity where it is cheap, sell it where it is dear, keep the spread.

Two things about such a trade are known in advance. First, it is magnificent while the gap is wide. Second, the gap must narrow — that is what arbitrage does to gaps. Indian wages rise, the n in X/n shrinks, and every firm in the industry has spent two decades building second moats — switching costs, scale, trust — to stand where the wage gap once stood. How well each firm built determines everything now. Which brings us back to the puzzle.

From cooking oil to code

Wipro's history is the most improbable in Indian technology. The company was founded in 1945 — before Independence, before computers — as Western India Vegetable Products, a maker of cooking oil in a small Maharashtra town. In 1966, its founder died suddenly, and his twenty-one-year-old son, Azim Premji, left Stanford mid-degree to run a vanaspati company. He proved to be one of the great capital allocators of his generation in one specific sense: he kept moving the firm's capital toward where the future was. Soaps, then hydraulic cylinders, then — when IBM exited India in the late 1970s — computers, and by the 1980s and 1990s, software services riding the same offshore wave as its Bengaluru neighbour Infosys.

For a while, Wipro was the wave. Around the turn of the century it stood among the most valuable companies in India, its founder briefly ranked among the richest men alive. That era also fixed the company's ownership character: the promoter family and its entities still hold 72.59% — and because much of that stake has been pledged to philanthropy, the invoice at the top of this chapter ultimately funds one of the largest charitable endowments on Earth. It is the noblest use of a services margin in world business. It does not, however, answer the puzzle of the margin itself.

The same trade, different hands

Now put Wipro's numbers beside the trade it shares with its peers, and let the gap speak.

Over the ten years to Mar 2026, Wipro compounded sales at 6% a year — against 9% at TCS, 11% at Infosys, 15% at HCL. In three of the last four years the top line barely moved at all: ₹90,488 crore in Mar 2023, then ₹89,760 crore, then ₹89,088 crore — two consecutive years of falling sales in a growing industry — before recovering to ₹92,624 crore. Profit compounded at 4% over the decade; ₹8,714 crore in Mar 2015 became ₹13,266 crore eleven years later. Operating margin runs at 19%, the lowest of the majors.

And the owners' return: 15.5% return on equity, 17.8% return on capital employed. Against TCS's 51.8% and 63.0%. Same invoice, same country, same clients. A third of the return.

Where did the spread leak? Not through the model — the model is the industry's. It leaked through execution and allocation, and the record names the channels. Wipro has bought growth repeatedly through large overseas acquisitions — consulting firms, design agencies, cloud specialists — paying full strategic prices for revenue that too often stopped growing once purchased; the borrowings line, at ₹20,291 crore the heaviest among the five majors (though still modest against reserves of ₹85,921 crore), is partly the residue. The corner office has been a revolving door by industry standards — insiders, then outside stars hired at global salaries, then insiders again — and each new occupant reorganised the firm, and each reorganisation cost a year of client focus. Meanwhile the operational metrics that decide services margins — utilisation, pricing discipline, large-deal win rates — stayed persistently a step behind the leaders. None of these wounds was fatal. All were self-inflicted. That is actually the optimistic reading: the gap to TCS is not geological; it is managerial.

Munger's post-mortem, and the decade of proof

Munger liked to say that he wanted to know where he would die, so he could avoid going there. Wipro is the instructive case for the whole industry, because it shows where the offshore model dies when merely average: not in bankruptcy — the model is too cash-generative for that — but in mediocrity with excellent cash flow. The switching costs that protect TCS protect Wipro too; clients do not leave, and so return on equity sits at 15.5% — above the cost of capital, below distinction, year after year. A moat exists. The castle inside it has been indifferently kept.

The incentive analysis is genuinely interesting. A 72.59% promoter holding devoted to philanthropy is a shareholder whose need is income, steady and rising — and the company's behaviour has visibly bent toward that need: dividend payout, erratic for a decade (as low as 5% of profit in three separate years), has jumped to 48% and then 87% in the last two years, and the stock now yields 6.29% — the highest of the five majors. For the endowment, that is design. An investor should simply see it clearly: this is a company increasingly run to distribute, not to compound.

What must management do this decade? The list is unusually concrete, because the benchmark walks around in plain sight. First, stop the revolving door — pick leadership and hold it for a decade; every services franchise that recovered (and the industry has examples) did it under boringly stable command. Second, declare an acquisition fast: Wipro's own record shows a rupee spent on integration discipline returns more than a rupee spent on the next marquee purchase. Third, fix the engine metrics — utilisation, pricing, delivery quality — before strategy; in services, strategy is mostly the compound interest of operational habits. Fourth, choose three battlegrounds (the firm has genuine strength in engineering and design services) rather than skirmishing everywhere the leaders fight. And fifth, face the AI transition without the luxury of a fat margin cushion: at 19% operating margin, Wipro has the least room of the majors to discount its way through the repricing of the engineer-hour.

What the owner gets for thirteen times earnings

The Buffett exercise, honestly done. The whole company sells for ₹1,72,619 crore — 13.2 times last year's ₹13,266 crore of profit, the cheapest multiple among the five majors, with a 6.29% dividend yield paid in cash.

The quality behind the price: a genuinely durable revenue base (the clients stay), operating cash flow of ₹14,932 crore last year — above reported profit, as it has been in most years; a balance sheet with ₹85,921 crore of reserves; earnings per share of ₹12.58 that have grown faster than profit (₹6.58 eleven years ago) as buybacks retired shares. The register shows retail investors arriving in size as institutions hesitated — shareholder count rose from 25.9 lakh to 30.6 lakh in under three years, presumably drawn by the yield.

An owner should be clear-eyed about what compounding is available. A business earning 15.5% on equity and paying out most of it retains little and grows slowly — the arithmetic points to an outcome like a well-secured bond with a slowly rising coupon. There is a genuine option attached: if some future management closes even half the execution gap to its peers, the same invoices would yield dramatically more profit. Options like that, however, have been available on Wipro for two decades, unexercised.

Verdict: narrow moat — the industry's switching costs and Wipro's sixty-year client relationships are real; everything above the moat has underperformed the identical opportunity set for twenty years, and the proof is the arithmetic: 15.5% against 51.8% on the same trade.

Where this trade goes to die — or not

The structural risks, with Wipro's specific exposure noted:

The base of the pyramid burns first. Machine-written code most cheaply replaces exactly what the offshore model bills most of: junior-engineer hours doing routine work. The invoice at the top of this chapter is repriced first at its bottom rung. Every Indian major faces this; the firm with 19% margins and 6% growth has the least cushion to absorb it and the most to gain from automating its own delivery aggressively.
The closing gap. Indian wage inflation shrinks the founding arbitrage every year, industry-wide. For leaders, second moats took the load; Wipro's second moat is thinner.
Client-country politics. Visas, data localisation, onshoring mandates — the shared climate of the industry, in full.
Distribution outrunning reinvestment. A payout policy shaped by an endowment's income needs could, in a hard decade, starve exactly the retraining and platform investment the AI transition demands.
The mediocrity trap itself. The cruellest structural risk: cash flows strong enough to fund every mistake indefinitely. Businesses like this rarely die; they dilute decades.

The long view from the oil press

Wipro's own history is the best argument that its story is not finished. This is a firm that has already been a vegetable-oil mill, a soap maker, a hydraulics manufacturer, a computer builder, and a global services company — an eighty-year lesson that a business is a pool of capital and people, not a product. The coming decades will force that lesson on the entire IT services industry at once, as machines absorb the work the industry was built to sell and leave behind the parts that were always hardest: judgement, integration, accountability, trust.

The industry will survive that passage; the interesting question is which firms cross it as leaders. Wipro brings to the crossing a durable client base, a fortress of reserves, a yield that pays the owner to wait — and a twenty-year execution gap that is, unlike wage arbitrage, entirely within its own power to close. The invoice is the same as everyone's. The hands are what's being tested.

An Omaha Investments chapter. Educational material, not investment advice. Figures from Screener.in and NSE data via Angel One as of the date above.